Efficient Data Management through ML-Based Feature Extraction
E R Adwaith Krishna, M. Varundev, Sidharth Surendran, Lekha S. Nair · 2024
Multimedia data production, such as audio, video, and acoustic measurements in large volumes and speed, has approached the scale of big data in many sensor networks. Multimedia Sensor Networks (MSN) are popular because of their capability to manage and integrate the different domains of sensor outputs, their representations, and the encoding of these measurements. Studies on generic and specific models for detection and tracing events in a sensor network have been established by multiple models. At the same time, there is still the requirement to handle these measurements from the Complex Dataset perspective. In the current world scenario, sensor networks are not limited to producing high-quality and diverse sensor readings but also large volumes at high speed. The most intuitive methods of handling them focus only on the interested range of measurement or store-for-later use. The possibility of data analysis in such an information-rich big data environment can be improved. In this study, we propose a model, “Multilink” that helps to align the multimedia data generated by a sensor network to a standard easy-to-process feature space with the incorporation of Machine Learning centric architecture.